The European Defence Agency (EDA) has conducted an explosives detection test using unmanned aerial and ground systems in Belgium.
The demonstration was part of the Artificial Intelligence for Detection of Explosive Devices (AIDED program) to gauge the potential of AI/machine learning, robot fleets, and sensors for bomb disposal applications.
During the trial, one aerial and two ground drones simulated a fully autonomous search to detect improvised explosive devices, including unexploded ordnance. They each incorporated advanced sensors and maneuvers in rural and urban scenarios.
Technologies integrated with the vehicles include electromagnetic metal detectors with ground penetrating radar and laser-induced breakdown spectroscopy to fire beams and create plasma on the soil for analysis and identification.
The systems leveraged neural network algorithms trained for specific signal processing. This allowed different sensors to provide detection in a single area.
The drones’ robot navigation employed localization and mapping capabilities to sustain operation across GPS-contested environments.
‘Effective’ Assessment
The demonstration faced minor challenges, such as sensor sensitivity issues due to soil humidity, movement of carrier systems, drone sensitivity due to high temperatures, and movement precision.
Even so, the drones exhibited detect and avoid maneuvers to prevent collision, enabling the unmanned aerial vehicle to hover very close to the targets during missions.
The systems achieved a Technology Readiness Level of three to four by the end of the test, validating their maturation for future bomb detection missions and further developments, EDA noted.
After this test, the agency will facilitate a follow-up project to evaluate the systems for hidden threat detection. Six partners from five EU countries will lead the initiative in four years.
“Overall the demonstration… showed how artificial intelligence can be used effectively in several aspects in an unmanned systems operational environment such as mission planning, self navigation, team working and last but not least explosive devices detection, tasks that have been performed autonomously with minimal human intervention,” EDA said.
